OpenCV 4 with Python Blueprints - Second Edition

Learning about grayscale features

The easiest feature to extract is probably the grayscale value of each pixel. Usually, grayscale values are not very indicative of the data they describe, but we will include them here for illustrative purposes (that is, to achieve baseline performance).

For each image in the input set, we are going to perform the following steps:

1. Resize all images to have the same (usually smaller) size. We use scale_size=(32, 32) to make sure we don't make the images too small. At the same time, we want our data to be small enough to work on our personal computer. We can do this with the following code:

resized_images = (cv2.resize(x, scale_size) for x in data)

2. Convert the image to grayscale (values are still in 0-255 ...